Strategy Verdict

Tag: Bitcoin

  • Bollinger Band Strategy Backtest: the 60-77% Win Rate That Loses on Every Coin

    Bollinger Band Strategy Backtest: the 60-77% Win Rate That Loses on Every Coin

    Bollinger Bands are the first indicator almost everyone learns, and the pitch is irresistibly simple: price is “cheap” at the lower band and “expensive” at the upper band, and it always snaps back to the mean. So you buy the lower band, sell the upper band, and collect. It even feels right — the win rate is high, the little green wins pile up. We ran the classic band-reversion through the full 7-Gate Protocol on 2 years, 5 coins and 5 timeframes. It is a reject, and it’s a perfect case study in why win rate is the most dangerous number in trading.

    Gate 00 — Fidelity (the easy one)

    No machine learning here. A Bollinger Band is a 20-period simple moving average plus/minus two population standard deviations — identical to TradingView’s ta.bb. There is nothing to reimplement incorrectly. The interesting question isn’t whether the bands are right; it’s whether reverting to them makes money.

    The Exact Rules

    • Bands: SMA(20) ± 2× standard deviation (the universal default)
    • Entry: go long when the close is below the lower band; go short when the close is above the upper band
    • Exit: close the position when price reverts to the middle band (the SMA)
    • Execution: signal on the bar close, enter at that close; costs 0.06%/side; data Binance spot Jul 2024–Jul 2026, BTC/ETH/SOL/BNB/XRP, 5m–1D; benchmark buy & hold
    Bollinger band reversion 7-gate verdict tearsheet reject, high win rate but loses on all coins

    The win-rate trap

    Bollinger reversion five coins 4H, 61 to 77 percent win rate yet loses on all five, XRP -98 percent

    Look at the win rates: 61% to 77% across all five coins. Textbook “high probability.” And yet it loses money on all five. The worst is XRP: a 67.5% win rate and a −98% return — nearly a wipeout. This is the whole lesson of mean reversion: you win often and small (price nudges back to the mean), then lose rarely and enormously (a trend runs and never reverts). A high win rate tells you how often you win, not how much — and here the maths is upside down.

    Why it bleeds: it fights every trend

    Bollinger band reversion on BTC 4H fights every trend, shorts breaks above the band and longs drops below

    By construction, band-reversion does the exact opposite of trend-following. Every time price breaks above the upper band — the start of many real rallies — it goes short. Every time price craters below the lower band — the middle of many real crashes — it goes long, catching the falling knife. In a ranging market that’s fine. Crypto over 2024–26 was anything but ranging, and the strategy spent two years standing in front of freight trains.

    Gate 06a — Every Timeframe

    Bollinger reversion net return by timeframe, self-destructs on lower timeframes, only daily positive

    On the timeframes the “scalping” clips love, it detonates: 5-minute bars → −100% (7,600 trades, account gone), 15m → −93%, 1h → −60%, 4h → −35%. Only the daily squeaks out +18% (30 trades, one symbol) — the familiar pattern where a strategy only survives where it barely trades.

    Gate 02 — Friction

    Bollinger reversion BTC 4H negative even at zero fees, no edge

    Here’s the tell that separates this from a merely fee-sensitive strategy: on BTC 4H it is negative even with zero fees (−21%). There is no gross edge for costs to erode — the reversion premise itself is a loser in a trending asset. Fees just deepen the hole.

    Gate 05 — Parameter Robustness

    Bollinger length by standard deviation parameter sweep, only 2 of 16 positive

    Maybe a different length or band width rescues it? We swept SMA length (10–50) against band width (1.5–3.0 SD): 2 of 16 combinations positive. There is no robust setting — only a couple of lucky cells.

    Gate 04 — Out-of-Sample

    Bollinger reversion out of sample 4H mostly losses no persistent edge

    Split 18 months in / 6 months out: mostly losses, only 2 of 5 coins positive out-of-sample, profit factors flipping across the boundary. Nothing stable to carry forward.

    The Verdict: REJECT

    • Gate 0 — Indicator fidelity — pass (standard SMA(20) ± 2 standard deviations, identical to TradingView’s ta.bb)
    • Gate 1 — Sanity — pass (signal on the closed bar, no look-ahead)
    • Gate 2 — Friction — fail (negative even at zero fees on 4H — there is no gross edge to erode)
    • Gate 3 — Yearly consistency — fail (unstable; it only survives in ranging stretches, not trending years)
    • Gate 4 — Out-of-sample — fail (2/5 positive out-of-sample)
    • Gate 5 — Robustness — fail (only 2 of 16 length × band-width cells positive)
    • Gate 6 — Multi-market — fail (loses money on all 5 coins; XRP −98% at a 67% win rate)
    • Gate 7 — vs Buy & Hold — fail (a 61–77% win rate and still behind simply holding)

    “Buy the lower band, sell the upper band” has a 60–77% win rate and loses money on every coin we tested. The high win rate is exactly the trap: it hides the rare, ruinous losses that happen when a trend refuses to revert. Mean reversion is not wrong as a concept — but naked band-reversion, with no regime filter, in a trending market, is a slow-motion blow-up. XRP: 67% winners, −98% equity.

    FAQ

    Doesn’t it work with an ADX / regime filter?
    Possibly — mean reversion is meant for ranging markets, and an ADX gate to sit out trends is the standard fix. But that is a different, more complex strategy, and choosing the filter after seeing the results is curve-fitting. The point of this test is the version that’s actually taught to beginners: bands only.

    How can a 77% win rate lose money?
    Because average win << average loss. You bank many tiny reversions and occasionally hold a losing position all the way through a trend. Win rate without payoff ratio is meaningless.

    Can I replicate this?
    Yes — SMA(20) ± 2SD, public Binance data, 0.06%/side. Every number falls out of those inputs.

    See also: RSI 30/70 (the other high-win-rate mean-reversion trap), Lorentzian Classification, UT Bot, and VWAP pullback (our one conditional pass).

    Disclaimer: educational research, not financial advice. Past performance does not guarantee future results. Never trade money you cannot afford to lose.

  • UT Bot Alerts Backtest: ‘5-Minute Easy Profits’ on 5 Coins and 5 Timeframes

    UT Bot Alerts Backtest: ‘5-Minute Easy Profits’ on 5 Coins and 5 Timeframes

    UT Bot Alerts is one of the most-screenshotted indicators on trading YouTube. “UT Bot makes crazy profits.” “The only indicator you need.” “5-minute easy money.” It is an ATR trailing-stop that paints a green “Buy” when price crosses above the stop and a red “Sell” when it crosses below — a clean, hypnotic flip-flop that looks unbeatable in a cherry-picked clip.

    We ran it through the full 7-Gate Protocol: 2 years of Binance data, 5 liquid coins, 5 timeframes from 5-minute to daily, a parameter sweep, an out-of-sample split, and two popular variants (Heikin-Ashi, EMA-200 filter). The verdict is reject — but the reason is genuinely interesting, and this time we begin by doing something most “backtest” videos never do: proving our indicator is bit-for-bit the real one.

    Gate 00 — First, we proved the indicator is real

    A backtest is worthless if the indicator underneath it doesn’t match what you actually see on TradingView. A single flipped comparison and your “short” becomes a “long” — the numbers look real but describe a strategy nobody trades. So before judging UT Bot, we reimplemented the canonical UT Bot Alerts Pine Script line-for-line and verified it three independent ways:

    • ATR match: our ATR equals TradingView’s ta.atr (Wilder RMA of true range) to 0.00 over all 4,380 bars — not “close”, identical.
    • Trailing-stop recurrence: the stop, recomputed by a second independent method, matches the engine to 0.00; and the stop never loosens mid-trend (0 ratchet violations) — the defining property of UT Bot.
    • No silent long/short inversion: across every timeframe and both close and Heikin-Ashi modes, every Buy coincides exactly with a flip to long and every Sell with a flip to short — zero same-bar contradictions, zero “wrong-side” signals. The direction is structurally incapable of inverting.

    Only after that check did we let it trade. Every number below rests on an indicator we can prove is the real thing.

    UT Bot indicator fidelity scorecard: ATR 0.00 diff over 4380 bars, trailing stop 0 diff, no long/short inversion, verification pass

    The Exact Rules

    • Indicator: UT Bot Alerts, default Key Value = 1, ATR Period = 10, source = close
    • Signal: price crosses above the ATR trailing stop → go long; crosses below → go short. It is always in the market (stop-and-reverse)
    • Execution: signal on the bar close, enter at that close — no look-ahead
    • Costs: 0.06% per side (fees + slippage), on every entry and exit
    • Data: Binance spot, Jul 2024 – Jul 2026, BTC/ETH/SOL/BNB/XRP, 5m–1D
    • Benchmark: buy & hold over the identical window
    UT Bot Alerts 7-gate verdict tearsheet, reject, BTC 2 years, key 1 ATR 10

    Gate 06a — Every Timeframe (the whole story)

    UT Bot net return by timeframe, 5m and 15m minus 100 percent, only daily positive, BTC

    Here is the punchline in one chart. UT Bot is sold on low timeframes — the “5-minute easy profits” clips — and that is precisely where it detonates. On BTC 5-minute bars it takes 31,242 trades and returns −100%: the account is gone. 15m: −100%. 1H: −92%. 4H: −27%. The only timeframe that stays alive is the daily (+46%, 76 trades) — the one timeframe nobody makes hype videos about. The pattern is monotonic: the faster you trade this thing, the faster it kills you.

    UT Bot on BTC daily chart, 77 signals in 2 years vs 31000 on 5-minute, trailing stop with buy sell markers

    The same indicator, same rules, on the daily — about 77 signals in two years instead of tens of thousands. That, and nothing cleverer, is why the daily survives fees while every intraday chart does not.

    Gate 02 — Friction (why it dies)

    UT Bot BTC total return vs fee per side, 4H dies at realistic fees, daily fee robust

    The mechanism is not mysterious. UT Bot’s gross edge is razor-thin — a profit factor of just 1.10 on 4H. With zero fees the 4H strategy makes +27%; at a realistic 0.06% per side it flips to −27%; at 0.11% it’s −54%. The entire “edge” is smaller than the transaction cost of harvesting it. The daily timeframe survives for one reason only: it trades ~75 times instead of 31,000, so friction barely touches it.

    Gate 03 — Yearly Consistency

    UT Bot yearly returns BTC, 4H loses every year, daily lives on 2025 alone

    On 4H, UT Bot loses money every single year (−14%, −11%, −5%). The daily’s headline +46% is not steady income either — it lives almost entirely in 2025 (+49%), with 2024 slightly negative and 2026 flat. One good year is not an edge; it’s a sample of one.

    Gate 06b — Five Coins

    UT Bot daily on five coins vs buy and hold, wins by shorting alt crashes, two of five lose

    On 4H, only 1 of 5 coins is profitable. On the daily it looks better — 3 of 5 positive and all three beat buy & hold — but read how:

    Coin (1D) UT Bot Buy & Hold Why
    BTC +46% +13% beat B&H in an up market
    ETH +11% −41% profited by shorting the crash
    SOL +175% −42% profited by shorting the crash
    BNB −22% +13% whipsawed, lost
    XRP −12% +158% missed the rally, lost

    The daily wins come from UT Bot’s short side catching the 2024–25 altcoin bear — something any long/short trend-follower would have done. It is not a UT-Bot-specific edge; it’s generic trend exposure, and it still failed on 2 of 5 coins with drawdowns of 50–80%.

    Gate 04 — Out-of-Sample

    UT Bot out of sample 4H, XRP in-sample plus 460 percent collapses to plus 4 percent

    Split each coin 18 months in-sample / 6 months out. The tell is XRP: a monstrous +460% in-sample collapses to +4% out-of-sample — the signature of a curve that fit noise. Only 2 of 5 coins are positive out-of-sample. Whatever looked like an edge does not survive contact with unseen data.

    Gate 05 — Parameter Robustness

    UT Bot key value by ATR period parameter sweep BTC 4H, only 3 of 20 cells positive

    We swept Key Value (1–3) against ATR Period (5–20) on BTC 4H: only 3 of 20 combinations are positive, and the profitable islands are scattered, not clustered. A real edge is robust to its own knobs; UT Bot’s result is a coin toss over the parameter grid.

    Do the popular variants save it?

    No. On BTC 4H: Heikin-Ashi source makes it worse (−48% vs the base −27%). An EMA-200 regime filter — only longing above the 200-EMA, shorting below — helps but stays negative (−17%). The combination is −30%. None of the “just add this one thing” fixes crosses zero.

    UT Bot BTC equity curve 4H vs 1D vs buy and hold, only daily stays above water

    Monte Carlo — Is the Daily’s +46% Just Luck?

    UT Bot Monte Carlo BTC daily 2000 bootstrap paths, median +46 percent, 71 percent beat buy and hold, wide -35 to +214 band

    We resampled the daily strategy’s returns into 2,000 block-bootstrap paths. Here the daily earns a little credit: the median path is +46% — the real result sits right on the median, not out on a lucky tail — and 71% of paths beat buy & hold. But the 5th–95th percentile band runs from −35% to +214%: even on its one survivable timeframe, UT Bot carries a real chance of a deep loss, and the fat right tail shows a handful of big trend trades doing most of the work. Dependable edges have tight distributions; this one does not.

    The Verdict: REJECT

    • Gate 0 — Indicator fidelity — pass (ATR and trailing stop match TradingView to 0.00; every Buy is a flip to long and every Sell a flip to short — the direction cannot invert)
    • Gate 1 — Sanity — pass (signal on the closed bar, entry at that close, no look-ahead)
    • Gate 2 — Friction — fail (+27% gross → −27% net on 4H; 5m and 15m go to −100%)
    • Gate 3 — Yearly consistency — fail (4H loses money every year; the daily’s +46% is almost entirely 2025)
    • Gate 4 — Out-of-sample — fail (2/5 positive; XRP +460% in-sample collapses to +4% out)
    • Gate 5 — Robustness — fail (only 3 of 20 key-value × ATR cells positive)
    • Gate 6 — Multi-market — fail (4H 1/5 coins; daily 3/5, and only via generic short exposure)
    • Gate 7 — vs Buy & Hold — mixed (beats it only on the daily, and only on 3 of 5 coins)

    UT Bot Alerts is marketed on exactly the timeframes where its constant flipping self-destructs on fees. Move it up to the daily and it stops bleeding — but that is the opposite of the “5-minute easy profits” pitch, and even there it’s a 3-of-5 coin toss with 30–80% drawdowns that mostly just shorted a bear market. The ATR trailing stop is a legitimate concept; UT Bot as sold is not a strategy.

    FAQ

    The videos show huge wins — are they lying?
    Usually they show gross returns (no fees) on a single lucky window and timeframe. Add real costs and test it out-of-sample on multiple coins and the picture inverts. That gap is the entire business model of hype channels.

    What about the Heikin-Ashi version everyone raves about?
    We tested it. On BTC 4H it was worse than the standard version (−48%). Smoother candles delay the flips but don’t create an edge.

    So the daily “+46%” means it works?
    On one coin, in one year, with a 31% drawdown and a tiny 76-trade sample. On the other four coins the daily is 2 losses and two wins that are really just “short the crash.” That is not a dependable edge.

    Can I replicate this?
    Yes — the rules above are complete, the data is public Binance OHLCV, fees 0.06%/side. We even proved the indicator matches TradingView to the decimal before running a single trade.

    See also: Triple SuperTrend (also −100% on 5m), the Golden Cross (loses to holding on all 5 coins), RSI 30/70 (died at gate 2), and VWAP pullback (our first conditional pass).

    Disclaimer: educational research, not financial advice. Past performance does not guarantee future results. Never trade money you cannot afford to lose.

  • Golden Cross Strategy: Does It Beat Buy & Hold? 9 Years, 5 Coins, 5 Timeframes

    Golden Cross Strategy: Does It Beat Buy & Hold? 9 Years, 5 Coins, 5 Timeframes

    The golden cross — a short moving average crossing above a long one — is the most famous signal retail traders know. YouTube sells it by the million views: “buy the golden cross, ride the trend.” So does it actually beat simply holding?

    We ran it through the full 7-Gate Protocol: 9 years of daily data, 5 liquid coins, 5 timeframes down to 5-minute bars, a parameter sweep, and a 2,000-path Monte Carlo. It failed — and not narrowly. Here is the complete autopsy.

    The Exact Rules

    • Signal: fast MA crosses above slow MA (default 50/200) → go long; death cross → flat (cash)
    • Position: spot, long-only, no leverage
    • Execution: signal on the daily close, enter the next bar — no look-ahead
    • Costs: 0.11% per side (fees + slippage), applied on every entry and exit
    • Data: Binance daily, Aug 2017 – Jul 2026 (plus intraday for the timeframe test)
    • Benchmark: buy & hold over the identical window

    The Baseline: BTC, 50/200

    Golden cross 50/200 seven-gate verification tearsheet, BTC 2017-2026, verdict reject

    +334% total return sounds great — until you see buy & hold did +1,362% over the same window. Sharpe 0.59 vs 0.79. Calmar 0.27. The strategy’s only real contribution is a smaller drawdown (−66.8% vs −83.2%). It doesn’t make you more; it loses you less. That is a risk tool, not a profit engine — and it is the opposite of what the golden-cross evangelists promise.

    Gate 03 — Yearly & Monthly

    Golden cross vs buy and hold yearly returns, BTC

    The pattern is unmistakable: the golden cross only wins in crash years — 2018 (−40% vs −73%), 2022 (−7% vs −64%). In every bull year it lags badly (2020: +109% vs +302%; 2021: +18% vs +60%). It is a drawdown filter wearing a profit strategy’s clothes.

    Golden cross monthly returns heatmap by year

    Monthly, there is no seasonality — long flat stretches (cash) punctuated by the occasional green month. You don’t get paid on a schedule; you get paid only when a big trend shows up.

    Gate 04 — Out-of-Sample

    Split the history in half. First half (2017–2022, an explosive bull): golden cross +133% vs buy & hold +759% — a rout. Second half (2022–2026, chop and decline): +86% (−37% DD) vs +69% (−67% DD) — here it wins, mostly on drawdown. The edge is regime-dependent, not stable.

    Gate 05 — Parameter Robustness (the killer)

    Golden cross CAGR by moving average pair, parameter robustness heatmap

    “Golden cross” means 50/200 to almost everyone. But CAGR swings wildly with the MA pair: the best combination reaches ~52% CAGR while the famous 50/200 sits near the bottom at ~18%. On ETH the gap is extreme — a 10/30 cross returns ~59% CAGR versus ~15% for 50/200. Most of the “performance” is simply which moving average you happened to pick. That is not an edge; that is luck, and the crowd picked one of the worst cells.

    Gate 06a — Every Timeframe (the scalper’s grave)

    Golden cross across timeframes, gross vs net after fees, friction destroys low timeframes

    Does it work intraday, the way the “golden cross scalping” videos claim? We ran the identical 50/200 cross on 5m, 15m, 1h, 4h and daily. The result is brutal: on 5-minute bars it is +74% gross but −71% net once you pay for 1,642 trades. 15m: +24% → −31%. The lower the timeframe, the faster friction eats it alive. 4H is the least-bad (+56% net) yet still loses to buy & hold (+74%) over the same window. Golden-cross scalping is dead on arrival — the fees alone bury it.

    Gate 06b — Five Liquid Coins

    Golden cross vs buy and hold on five liquid coins BTC ETH SOL XRP BNB
    Symbol Golden Cross Buy & Hold Strategy Max DD Verdict
    BTC +334% +1,362% −66.8% loses to holding
    ETH +245% +481% −79.4% loses to holding
    SOL +922% +2,272% −78.4% loses to holding
    XRP −65% +23% −91.4% loses money
    BNB +1,740% +36,044% −77.3% loses to holding

    Five of the most liquid coins in crypto. The golden cross loses to buy & hold on every single one. On XRP it doesn’t merely underperform — it loses money (−65% while holding made +23%), whipsawed to death by a coin that mostly went sideways. Note BNB: even against a 361× monster, the golden cross returned “only” +1,740% — underperforming by a factor of twenty. The stronger the trend, the more it leaves on the table.

    Gate 07 — Monte Carlo

    Golden cross Monte Carlo 2000 bootstrap paths terminal return distribution

    2,000 block-bootstrap paths of the daily returns. The median outcome is +375% — close to the actual +334%, and still far below buy & hold’s +1,362%. The 5th–95th percentile band is enormous (−61% to +5,123%), confirming heavy path-dependence. Translation: even luck doesn’t rescue it.

    Take-profit to stop-loss — and why it cannot help

    Could a fixed target beat holding to the death cross? We swept a fixed stop (3×ATR) and target on the daily 50/200 crosses, 2017–2026.

    TP : SL Trades Win rate Profit factor Net return
    1 : 0.5 8 62.5% 0.75 −12%
    1 : 1 8 50.0% 1.01 −5%
    1 : 1.5 8 50.0% 1.52 +16%
    1 : 2 8 50.0% 2.03 +40%
    1 : 2.5 8 50.0% 2.55 +69%
    1 : 3 8 50.0% 3.06 +101%

    Two problems leap out. First, the sample: a 50/200 golden cross fires just 8 times in nine years, so every figure here rests on eight trades — statistically meaningless, however tidy the rising profit factor looks. Second, even the best fixed target (+101% at 1:3) is far below simply holding to the death cross (+334%), and further still below buy-and-hold (+1362%). A take-profit on a trend-following cross only caps the rare winners that justify it. TP:SL cannot rescue a signal already beaten by doing nothing.

    The Verdict: REJECT

    Adding up the gates:

    • Gate 0 — Indicator fidelity — pass (standard 50/200 simple-moving-average cross)
    • Gate 1 — Sanity — pass (signal on the closed bar, entry next bar, no look-ahead)
    • Gate 2 — Friction — fail intraday (5m/15m flip to losses purely on fees)
    • Gate 3 — Yearly consistency — fail (wins only in crash years, lags every bull)
    • Gate 4 — Out-of-sample — partial (regime-dependent)
    • Gate 5 — Robustness — fail (result hinges on the MA pair; 50/200 is near-worst)
    • Gate 6 — Multi-market — fail (loses to holding on all 5 coins)
    • Gate 7 — vs Buy & Hold — fail (behind on total return and Sharpe)

    “Buy the golden cross and profit” is, on the data, false. Its only value is drawdown reduction — and even that depends on a moving-average pair you got lucky with. As a standalone profit strategy it is a myth. As a defensive trend filter bolted onto a real system, maybe. Nothing more.

    That distinction is the whole point of this site. One YouTube title says “The golden cross made 922% on Solana!” Another says “I tested the golden cross and lost money!” Both are technically true (SOL and XRP, right here). Neither is the truth.

    FAQ

    Then why does everyone still use it?
    Because in crash years it looks like genius, and almost nobody backtests the bull years it quietly lags. Availability bias does the rest.

    Isn’t a smaller drawdown worth it?
    Only if you value risk reduction over returns and accept the parameter fragility. A plain “exit below the 200-day” achieves similar defense with far less overfitting risk.

    Does it work on stocks or forex?
    Different assets, different result. This test is crypto spot, daily and intraday. We only publish what we measured.

    Can I replicate this?
    Yes — the rules above are complete, the data is public Binance OHLCV, fees 0.11%/side. Every number in this article falls out of those inputs.

    See also: the RSI 30/70 strategy (died at gate 2) and the VWAP trend-pullback strategy (the first to earn a conditional pass).

    Disclaimer: This is educational research, not financial advice. Past performance does not guarantee future results. Never trade money you cannot afford to lose.

  • VWAP Strategy Backtest: 5 Coins, 16 Parameter Sets, Full Out-of-Sample Test

    VWAP Strategy Backtest: 5 Coins, 16 Parameter Sets, Full Out-of-Sample Test

    Last time, the RSI 30/70 strategy died at gate 2 — fees ate it alive before we even got to the interesting questions. This strategy is different. It’s the first one to make it deep into the 7-Gate Protocol: five symbols, sixteen parameter sets, and a full out-of-sample split.

    It survived more gates than anything we’ve tested. It still didn’t survive all of them. Here’s the complete autopsy — including exactly where it works and where it dies.

    The Exact Rules

    • Timeframe: 4H candles
    • Trend filter: EMA(100) — longs only above it, shorts only below it
    • Entry: price pulls back and touches the daily VWAP, then closes back in the trend direction
    • Stop loss: fixed at entry ± 2.0 × ATR(14) — never trailed
    • Exit: close crossing back through the EMA(100) (trend over), or the stop
    • Fees: 0.06% per side, intrabar stop fills
    • Data: 2 years (July 2024 – July 2026), Binance public data

    One counterintuitive detail from our earlier testing: a trailing stop destroys this strategy (PF 0.76). Pullback entries get shaken out by noise. The fixed stop is not a preference — it’s the difference between profit and ruin.

    The Baseline: BTC, 4H

    VWAP trend pullback strategy backtest tear sheet BTCUSDT 4H with trade markers

    Look at that win rate: 22.7%. Three losses out of four trades — and it still made +25.5%, double Buy & Hold. This is the exact mirror image of the RSI lesson: average win +6.77%, average loss −1.45%. Win rate is a vanity metric. Payoff asymmetry is the business model.

    Gate 03 — Monthly Consistency

    VWAP strategy monthly returns heatmap by symbol

    Not pretty, not terrible. Long flat-to-red stretches punctuated by big green months — the classic trend-following profile. You don’t get paid monthly; you get paid when trends happen.

    Gate 04 — Out-of-Sample

    VWAP strategy in-sample vs out-of-sample returns by symbol

    We split the data: first 18 months in-sample, last 6 months untouched. 3 of 5 symbols stayed positive out-of-sample. BTC actually got better (PF 2.46 out-of-sample). SOL and XRP flipped negative. Partial pass — the edge doesn’t evaporate on unseen data, but it’s not universal either.

    Gate 05 — Parameter Robustness

    VWAP strategy parameter sweep heatmap EMA ATR robustness

    Sixteen combinations of EMA length (50–200) and ATR stop multiple (1.5–3.0): 13 of 16 positive. This is what a real edge looks like — it degrades gracefully when you wiggle the knobs. A curve-fit strategy shows one green cell in a sea of red.

    Gate 06 — Five Symbols, Same Rules

    VWAP strategy tested on five crypto symbols vs buy and hold
    Symbol Trades PF Total return Max DD Buy & Hold Verdict
    BTC 128 1.37 +25.5% 36.7% +12.1% beats holding
    ETH 128 1.51 +83.3% 41.8% −39.8% crushes holding
    SOL 133 1.15 +4.2% 51.1% −40.2% beats holding
    BNB 165 0.80 −50.8% 60.6% +16.8% fails
    XRP 148 1.68 +1.5% 75.9% +163.2% loses to holding

    This is why gate 06 exists. Test on BTC alone and you’d call it a winner. Test on BNB and you’d call it garbage. Both would be wrong: the edge is real on majors and absent elsewhere. Anyone selling you a strategy that “works on everything” hasn’t run this test.

    Every timeframe (why 4-hour is the whole story)

    A conditional pass comes with a condition, and for VWAP pullback it is the timeframe. The identical rules, run from 5-minute to daily:

    Timeframe Trades Win rate Profit factor Net return
    5-minute 9,124 11.7% 0.22 −100%
    15-minute 3,097 15.1% 0.37 −100%
    30-minute 1,442 18.4% 0.54 −93%
    1-hour 709 20.3% 0.71 −68%
    4-hour (baseline) 159 22.6% 1.20 +21%
    1-day 30 20.0% 0.73 −28%

    Every intraday interval loses — a total wipe on 5-minute — and the edge appears only on the 4-hour. Push to daily and the sample thins and it slips back to −28%. This is not a strategy you can run anywhere; it is a 4-hour phenomenon, which is precisely why the verdict is conditional.

    Take-profit to stop-loss (4-hour)

    Swapping the exit for a fixed stop (2×ATR) and a swept target:

    TP : SL Trades Win rate Profit factor Net return
    1 : 0.5 102 69.6% 0.93 −8%
    1 : 1 77 57.1% 1.15 +13%
    1 : 1.5 72 50.0% 1.29 +31%
    1 : 2 61 41.0% 1.16 +14%
    1 : 2.5 55 38.2% 1.29 +27%
    1 : 3 53 32.1% 1.11 +5%

    Unlike the rejects on this site, VWAP’s edge survives the exit sweep: for every reward-to-risk from 1:1 to 1:3 the profit factor holds above 1.0, peaking at 1.29. That resilience across exits is what separates a conditional pass from a rejection — though a very tight 1:0.5 target still tips it slightly negative.

    The Verdict: CONDITIONAL

    Adding up the gates:

    • Gate 0 — Indicator fidelity — pass (standard daily VWAP with an EMA trend filter and ATR stop)
    • Gate 1 — Sanity — pass (signals on closed bars, no look-ahead)
    • Gate 2 — Friction — pass (still profitable after 0.06%/side — the gate where RSI died)
    • Gate 3 — Yearly consistency — partial (trend-dependent, with long flat stretches)
    • Gate 4 — Out-of-sample — partial (3 of 5 positive)
    • Gate 5 — Robustness — pass (13 of 16 sweep cells positive)
    • Gate 6 — Multi-market — partial (works on majors; BNB fatal)
    • Gate 7 — vs Buy & Hold — partial (beats it on 3 of 5 coins)

    Our verdict stamp says CONDITIONAL, and now the conditions are precise:

    Majors only (BTC/ETH). Reduced position size — the 37–42% drawdowns are real. As a portfolio component, not a standalone system. Traded outside those conditions, expect the BNB outcome.

    That nuance is the whole point of this site. A YouTube title would say “this VWAP strategy made 83% on ETH!” A different YouTube title would say “I tested VWAP and lost 50%!” Both are technically true. Neither is the truth.

    FAQ

    Why does it fail on BNB?
    BNB spent much of the window in choppy, range-bound conditions where trend filters generate false regime signals. The strategy needs trends; BNB didn’t provide them.

    Would a different VWAP (weekly, anchored) help?
    In our earlier tests, weekly VWAP performed worse as a touch level. Volume-profile POC levels also degraded results — daily VWAP carried all the edge.

    What about the low win rate — can I handle 3 losses out of 4?
    That’s the real question. Statistically it works; psychologically most people abandon it during the losing streaks. That’s a you-parameter, not a strategy parameter.

    Can I replicate this?
    Yes — the rules above are complete, data is public Binance OHLCV, fees 0.06%/side. Every number in this article falls out of those inputs.

    Disclaimer: This is educational research, not financial advice. Past performance does not guarantee future results. Never trade money you cannot afford to lose.

  • RSI Strategy Backtest: I Tested the Famous 30/70 Rule on Bitcoin (It Lost 65%)

    RSI Strategy Backtest: I Tested the Famous 30/70 Rule on Bitcoin (It Lost 65%)

    Every trading YouTube channel eventually makes the same video: “Buy when RSI drops below 30, sell when it crosses 70.” It sounds logical. It looks great on cherry-picked charts. Some videos claim win rates of 80–90%.

    So I did what almost nobody does: I coded the exact rules and ran them on 2 years of real Bitcoin data — with real trading fees included.

    Spoiler: every variant lost money. One lost 65%. Here is the full breakdown, so you don’t have to pay for this lesson with your own account.

    The Exact Rules I Tested

    No vague “price action confirmation.” Rules a computer can execute:

    • Indicator: RSI(14), Wilder’s smoothing, 1-hour candles
    • Long entry: RSI crosses up through 30
    • Long exit: RSI crosses up through 70
    • Short entry (long+short variant): RSI crosses down through 70
    • Short exit: RSI crosses down through 30
    • Data: BTCUSDT, 17,500+ hourly candles (July 2024 – July 2026)
    • Fees: 0.06% per side (typical crypto futures taker fee)
    • Position size: 100% of equity per trade, starting from $10,000

    The Results

    RSI strategy backtest full tear sheet with long short trade markers bitcoin
    Variant Trades Win rate Profit factor Total return Max drawdown
    RSI 30/70 long+short 125 54.4% 0.75 −64.9% 73.1%
    RSI 30/70 long-only 62 58.1% 0.86 −27.1% 46.6%
    RSI 20/80 long+short 26 69.2% 1.05 −38.3% 73.3%
    Buy & Hold BTC — — — +12.7% —
    RSI strategy total returns after fees bar chart

    Read that table again. The strictest variant had a 69% win rate and still lost 38%. Meanwhile, doing absolutely nothing — just holding BTC — made +12.7%.

    Why a 69% Win Rate Still Loses Money

    This is the single most important lesson in this article.

    RSI mean-reversion produces many small wins and a few catastrophic losses. When you buy an oversold dip in a real downtrend, RSI doesn’t politely bounce back. It stays oversold while price keeps falling — and the strategy has no stop loss. One bad trend wipes out twenty small wins.

    RSI strategy 73 percent drawdown chart

    That’s what a 73% drawdown looks like. If you started with $10,000, at the worst point you had $2,700. Nobody keeps trading a system through that.

    The math that YouTube never shows:

    • Win rate is meaningless without payoff ratio. 69% wins × small size, 31% losses × huge size = net loss.
    • Fees compound brutally. 125 round trips × 0.12% ≈ 15% of your account gone to fees alone.
    • Buying dips fights the trend. In crypto, trends run further than RSI assumes.

    “But It Worked in That YouTube Video…”

    1. Cherry-picked windows. Any strategy looks amazing during the right 3 months. I tested a full 2-year window.
    2. No fees or slippage. Add 0.06% per side and high-frequency signals collapse.
    3. Hindsight entries. In live trading you get every RSI<30 signal, including the twenty that came before the bottom.
    RSI strategy backtest equity curve vs buy and hold bitcoin

    Same Rules, Every Timeframe: 5m to 1D

    “Maybe it just needs a lower timeframe.” I hear that every time a strategy fails. So the engine re-ran the identical rules on five timeframes — over 240,000 candles in total. Nobody gets to say I didn’t look.

    RSI strategy tested on 5m 15m 1h 4h 1d timeframes total returns
    Timeframe Trades Win rate Profit factor Total return Max drawdown
    5m 1,436 65.7% 0.99 −86.0% 88.2%
    15m 466 63.9% 1.07 −36.8% 56.1%
    1h 125 54.4% 0.75 −64.9% 73.1%
    4h 38 65.8% 0.93 −36.0% 64.0%
    1d 5 80.0% 1.13 −14.7% 66.5%

    Two things this table screams:

    • 5m is death by fees. A 65.7% win rate across 1,436 trades — and it still lost 86%. At 0.12% per round trip, the fees alone consumed more than the entire account. This is gate 02 in its purest form.
    • 1D hit an 80% win rate and still lost money. Five trades, four winners — and the single loser erased them all. Win rate tells you nothing about the size of the loss that’s coming.
    RSI strategy monthly returns heatmap by timeframe

    The monthly heatmap exposes the regime problem. In 2025 — an up-trending year — the 1h/4h/1d variants printed +22% to +39%. Then 2024 and 2026 took it all back:

    Timeframe 2024 (H2) 2025 2026 (H1)
    5m −24.9% −71.5% −34.7%
    15m −26.8% −3.5% −10.4%
    1h −64.1% +22.5% −20.2%
    4h −47.0% +38.9% −13.1%
    1d −41.4% +23.6% +17.8%

    A strategy that only works in one market regime isn’t a strategy — it’s a bet on the regime. That’s a gate 03 failure (yearly consistency), stacked on top of the gate 02 failure.

    Does a fixed take-profit save it? (TP : SL sweep)

    Maybe the RSI-70 exit is the problem. We replaced it with a fixed stop (2×ATR) and a fixed target, and swept the reward-to-risk from 1:0.5 to 1:3 on BTC 1H.

    TP : SL Trades Win rate Profit factor Net return
    1 : 0.5 254 64.6% 0.60 −48%
    1 : 1 218 42.7% 0.60 −60%
    1 : 1.5 210 34.3% 0.69 −56%
    1 : 2 203 27.6% 0.71 −56%
    1 : 2.5 197 23.4% 0.72 −56%
    1 : 3 190 20.5% 0.73 −56%

    It changes nothing. Every ratio loses 48–60% with a profit factor stuck between 0.60 and 0.73. A tight target lifts the win rate to 65%, but the losers are twice the size; a wide target trims the win rate to 20%. There is no exit that rescues a losing entry — the oversold bounce simply is not there to harvest.

    Does This Mean RSI Is Useless?

    No — it means RSI as a standalone entry signal is useless on crypto. Tools like RSI or VWAP only stop bleeding money when they’re subordinated to a trend filter with a hard stop loss — and even then they rarely beat a simple trend-following system. (Read the full VWAP backtest here — it made it much further through the gates.)

    The general rule: mean reversion without a stop loss is how accounts die slowly, then suddenly.

    The 7-Gate Scorecard

    • Gate 0 — Indicator fidelity — pass (standard Wilder RSI(14))
    • Gate 1 — Sanity — pass (signal on the closed bar, no look-ahead)
    • Gate 2 — Friction — fail (5-minute loses 86% on fees alone; negative on every timeframe)
    • Gate 3 — Yearly consistency — fail (positive only during 2025’s bull; loses in every other stretch)
    • Gate 4 — Out-of-sample — fail (2 of 5 coins positive out-of-sample)
    • Gate 5 — Robustness — fail (every take-profit ratio from 1:0.5 to 1:3 loses)
    • Gate 6 — Multi-market — fail (0 of 5 coins profitable on 4H; XRP −977% at an 83% win rate)
    • Gate 7 — vs Buy & Hold — fail (long+short −65%, long-only −27% vs BTC +12.7%)

    How I Validate Any Strategy (The 7-Gate Checklist)

    1. Code sanity — no lookahead bias, no repainting
    2. Friction — realistic fees and slippage included
    3. Yearly breakdown — profits every year, or one lucky year?
    4. Out-of-sample — does it survive data it wasn’t tuned on?
    5. Robustness — small parameter changes shouldn’t destroy it
    6. Multi-market — one coin’s fluke, or a general edge?
    7. Beats Buy & Hold — otherwise, why bother?

    The RSI 30/70 strategy fails gate 2 and never recovers. Most YouTube strategies die at the same gate.

    FAQ

    What RSI settings did you use?
    RSI(14), Wilder’s smoothing, 1H candles — the default in TradingView.

    Would a stop loss fix it?
    It reduces catastrophic losses but doesn’t create an edge. The entry itself is the problem.

    What data and code did you use?
    Public Binance OHLCV data and a Python backtester. The rules above are complete — you can replicate every number.

    Disclaimer: This is educational research, not financial advice. Past performance does not guarantee future results. Never trade money you cannot afford to lose.